Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Brief one-line description of what this skill does
.claude/skills/heliosarchitect-skill-name/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 62% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 12% | 0% |
Brief description of what this skill does and why it's useful.
REQUIRED_API_KEY - Description of what this key doesjq - JSON processorcurl - HTTP clientbash# Clone to your skills directory cd ~/.openclaw/workspace/skills/ git clone https://github.com/yourusername/your-skill.git skill-name
bashopenclaw skill install skill-name
Set required environment variables in your OpenClaw config or shell:
bashexport REQUIRED_API_KEY=your_key_here
Or add to ~/.openclaw/openclaw.json:
json{ "env": { "REQUIRED_API_KEY": "your_key_here" } }
bash./scripts/main_script.sh arg1 arg2
The skill is automatically available when loaded. OpenClaw will detect and offer it based on context.
See examples/ directory for common use cases.
SKILL.md - This file (skill documentation + frontmatter)scripts/ - Executable scripts and toolsreferences/ - Documentation, guides, linksassets/ - Images, data files, templatesexamples/ - Example usage and sample outputsbash# Run tests ./scripts/test.sh
Problem: Script fails with "command not found"
Problem: API authentication fails
REQUIRED_API_KEY is set correctlyMIT License - see LICENSE file
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 7,193 | 11,144 | +55% | 1 | 1 | 0% | 1,335 | 2,165 | +62% | 0 | 0 | — |
case-02 | fail→pass | 6,164 | 3,397 | -45% | 1 | 1 | 0% | 1,070 | 1,376 | +29% | 0 | 0 | — |
case-03 | fail→pass | 13,262 | 8,738 | -34% | 1 | 1 | 0% | 2,287 | 2,240 | -2% | 0 | 0 | — |
case-04 | pass→pass | 12,794 | 1,917 | -85% | 1 | 1 | 0% | 2,080 | 878 | -58% | 0 | 0 | — |
case-09 | pass→pass | 5,672 | 1,928 | -66% | 1 | 1 | 0% | 905 | 803 | -11% | 0 | 0 | — |
case-05 | pass→pass | 5,593 | 2,508 | -55% | 1 | 1 | 0% | 943 | 962 | +2% | 0 | 0 | — |
case-06 | fail→pass | 7,848 | 2,520 | -68% | 1 | 1 | 0% | 1,383 | 841 | -39% | 0 | 0 | — |
case-07 | fail→pass | 7,893 | 5,345 | -32% | 1 | 1 | 0% | 1,429 | 1,601 | +12% | 0 | 0 | — |
case-08 | pass→pass | 6,411 | 1,884 | -71% | 1 | 1 | 0% | 1,060 | 786 | -26% | 0 | 0 | — |
case-10 | fail→pass | 8,291 | 1,885 | -77% | 1 | 1 | 0% | 1,479 | 863 | -42% | 0 | 0 | — |
case-11 | pass→pass | 7,647 | 2,593 | -66% | 1 | 1 | 0% | 1,321 | 939 | -29% | 0 | 0 | — |
case-12 | pass→pass | 5,582 | 1,924 | -66% | 1 | 1 | 0% | 846 | 808 | -4% | 0 | 0 | — |
case-13 | fail→pass | 9,256 | 3,230 | -65% | 1 | 1 | 0% | 1,599 | 901 | -44% | 0 | 0 | — |
case-14 | pass→pass | 13,904 | 1,811 | -87% | 1 | 1 | 0% | 2,395 | 827 | -65% | 0 | 0 | — |
case-15 | fail→pass | 7,129 | 3,001 | -58% | 1 | 1 | 0% | 1,231 | 1,036 | -16% | 0 | 0 | — |
case-16 | fail→pass | 8,755 | 1,971 | -77% | 1 | 1 | 0% | 1,498 | 888 | -41% | 0 | 0 | — |
case-17 | fail→pass | 9,152 | 2,142 | -77% | 1 | 1 | 0% | 1,567 | 916 | -42% | 0 | 0 | — |
case-18 | pass→pass | 3,803 | 3,362 | -12% | 1 | 1 | 0% | 553 | 762 | +38% | 0 | 0 | — |
case-19 | pass→pass | 4,751 | 1,599 | -66% | 1 | 1 | 0% | 761 | 755 | -1% | 0 | 0 | — |
case-20 | pass→pass | 6,724 | 4,578 | -32% | 1 | 1 | 0% | 1,223 | 1,533 | +25% | 0 | 0 | — |
case-21 | pass→pass | 13,200 | 11,172 | -15% | 1 | 1 | 0% | 2,402 | 2,638 | +10% | 0 | 0 | — |
case-22 | pass→pass | 9,869 | 8,384 | -15% | 1 | 1 | 0% | 1,756 | 2,140 | +22% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +45 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.